On 12/13/2012 1:21 PM, Dwayne Reid wrote:
> Its going to cost us somewhere between $5K - $10K to get
> certification on this new design and I'd *really* not spend that kind
> of money, then find that I have a show-stopper problem relating to
> the PIC controller.  Unfortunately, this is one of those chicken and
> egg problems - I can't ship these heaters into their target
> environment without certification and I can't be certain that they
> won't fail until they are in their target environment (really crappy powe=
r).
>
> Many thanks!
>
> dwayne
>
I have no information for you about the reliability of modern PICs. But=20
the essence of your question is exactly the kind of thing that keeps me=20
awake at nights.

Having had a similar problem once, I built pic-based board that would do=20
various things to the power supply of a unit under test. The test unit=20
had software that tested as much of it's normal functionality as=20
possible.  When the controller pic would do something to the power=20
supply (jerk it down half way, etc.), the unit under test would be=20
monitored and discrepancies noted.  There never were any for the=20
particular situation.  I probably didn't try hard enough.  If I were in=20
your shoes, I would think about getting a setup that could manipulate=20
the targets power supply in many different ways and have a monitor that=20
would log the results.  This is no absolute guarantee that the target=20
system is reliable; it simply helps find ones that are particularly=20
unreliable.
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.